[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20210924124330.173774576@linuxfoundation.org>
Date: Fri, 24 Sep 2021 14:44:00 +0200
From: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To: linux-kernel@...r.kernel.org
Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
stable@...r.kernel.org, Li Huafei <lihuafei1@...wei.com>
Subject: [PATCH 4.19 06/34] tracing/kprobe: Fix kprobe_on_func_entry() modification
From: Li Huafei <lihuafei1@...wei.com>
The commit 960434acef37 ("tracing/kprobe: Fix to support kretprobe
events on unloaded modules") backport from v5.11, which modifies the
return value of kprobe_on_func_entry(). However, there is no adaptation
modification in create_trace_kprobe(), resulting in the exact opposite
behavior. Now we need to return an error immediately only if
kprobe_on_func_entry() returns -EINVAL.
Fixes: 960434acef37 ("tracing/kprobe: Fix to support kretprobe events on unloaded modules")
Signed-off-by: Li Huafei <lihuafei1@...wei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
kernel/trace/trace_kprobe.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/kernel/trace/trace_kprobe.c
+++ b/kernel/trace/trace_kprobe.c
@@ -836,8 +836,9 @@ static int create_trace_kprobe(int argc,
pr_info("Failed to parse either an address or a symbol.\n");
return ret;
}
+ /* Defer the ENOENT case until register kprobe */
if (offset && is_return &&
- !kprobe_on_func_entry(NULL, symbol, offset)) {
+ kprobe_on_func_entry(NULL, symbol, offset) == -EINVAL) {
pr_info("Given offset is not valid for return probe.\n");
return -EINVAL;
}
Powered by blists - more mailing lists